NIR autofluorescence allows for pituitary gland detection during surgery: the first evidence from microscopic studies and in vivo measurements

This study demonstrates that near-infrared autofluorescence (NIRAF), driven by secretory granules, can effectively distinguish normal pituitary tissue from tumors during surgery, offering a promising label-free method for enhancing gland preservation.

The Gist

Imagine you are a surgeon performing a delicate operation deep inside a patient's nose to remove a brain tumor. Your goal is to take out the bad tumor while leaving the healthy "control center" (the pituitary gland) completely untouched. The problem? Under a normal white light microscope, the healthy gland and the tumor look almost identical. They are like two different types of dough that have been kneaded together; it's hard to tell where one ends and the other begins. If you cut too much, the patient loses their hormonal balance. If you cut too little, the tumor grows back.

This paper presents a brilliant new "flashlight" that solves this problem without using any dyes or chemicals.

The "Glow-in-the-Dark" Secret

The researchers discovered that the healthy pituitary gland has a natural, built-in glow in the near-infrared spectrum (a type of light invisible to the human eye but visible to special cameras). Think of it like a firefly.

• The Healthy Gland: It is packed with tiny, glowing fireflies (scientists call these "secretory granules"). When you shine a specific red light on it, these fireflies light up brightly.
• The Tumor: The tumor cells are like a factory that stopped making fireflies. They have very few of these glowing granules. When you shine the same red light on the tumor, it stays dark.

The Detective Work (Microscope Studies)

Before trying this on real patients, the team played detective in the lab. They took samples of tissue and looked at them under a super-powerful microscope that could see the "colors" of the light coming from the cells.

They found that the healthy gland was a bustling city full of glowing granules, while the tumor was a quiet, dark neighborhood. They realized that the "glow" wasn't coming from the cell walls or the blood, but specifically from these tiny granules inside the cells. The more granules, the brighter the glow.

The Surgery (The Real Test)

The team then took this discovery into the operating room. They used a special, sterile fiber-optic probe (like a tiny, glowing wand) that the surgeon could gently touch to the tissue during the operation.

1. The Setup: The surgeon shines a red laser (650 nm) onto the tissue.
2. The Reaction: The healthy pituitary gland immediately glows brightly in the infrared spectrum. The tumor and surrounding tissues (like bone or the lining of the brain) remain dim.
3. The Result: The surgeon can now "see" the boundary between the healthy gland and the tumor in real-time. It's like having a GPS for your scalpel.

How Well Did It Work?

The results were almost perfect.

• They tested this on 27 different patients.
• In every single case, the healthy gland glowed brighter than the tumor.
• They used a statistical test (called an ROC-AUC) to measure accuracy, and the score was 0.98 out of 1.0. In the world of medical tests, this is like getting an A+ on every single exam. It means the method is incredibly reliable at telling the difference between the "good" tissue and the "bad" tissue.

Why This Matters

Currently, surgeons rely on their eyes and experience, which can be tricky when the tumor is invasive or the tissue is scarred. This new method is label-free, meaning they don't need to inject the patient with expensive, potentially toxic dyes (like the ones used in other surgeries). The gland glows on its own.

The Big Picture:
Think of this technology as giving the surgeon X-ray vision for the pituitary gland. Instead of guessing where the healthy tissue ends, they can see a bright, glowing beacon that says, "Stop! This is the healthy gland!" This allows them to remove the tumor more completely while preserving the patient's hormonal health, leading to better recovery and fewer complications.

In short, the researchers found that the pituitary gland has a natural "superpower" (a glow) that tumors lack, and they built a tool to help surgeons see that superpower in real-time.

Technical Summary

1. Problem Statement

The primary challenge in endocrine neurosurgery, specifically for Pituitary Neuroendocrine Tumors (PitNETs), is the intraoperative discrimination between normal pituitary tissue and tumor tissue.

• Clinical Context: Gross-total resection is achieved in only ~80% of cases, with high regrowth rates (up to 75%) after incomplete removal.
• Limitations of Current Methods:
  • Visual Inspection: Differentiating tumor from normal tissue is difficult, especially in fibrotic or infiltrative tumors, leading to inadvertent damage of healthy gland tissue and subsequent endocrine disorders.
  • Exogenous Agents: Fluorescent agents like Indocyanine Green (ICG), 5-ALA, and targeted probes (e.g., OTL38, bevacizumab-800CW) have limitations including sensitivity to artifacts (blood, mucosa), variable tumor-to-normal contrast, complex dosing protocols, and dependence on specific receptor expression profiles.
• Goal: Develop a label-free intraoperative guidance method to distinguish normal pituitary tissue from PitNETs and surrounding structures to enhance gland preservation.

2. Methodology

The study employed a multi-stage approach combining ex vivo microscopic analysis with in vivo intraoperative measurements.

A. Ex Vivo Microscopic Analysis (Mechanism Verification)

• Samples: 40 human pituitary specimens (22 somatotroph, 11 corticotroph, 4 non-functioning, 2 prolactinomas, 1 thyrotroph) and 9 confirmed normal pituitary samples.
• Imaging Techniques:
  • Confocal Laser Scanning Microscopy (CLSM): Used Olympus FV3000 with 405, 488, 561, and 640 nm excitation.
  • Spectral Imaging: Captured emission spectra (420–800 nm) to identify fluorophores.
  • Reflectance Imaging: Used 640 nm to visualize tissue morphology.
• Image Analysis: Regions of Interest (ROIs) were segmented into three classes: Secretory Granules, Collagen, and Cytoplasm. Quantitative analysis compared the area fraction and spectral properties of these ROIs between normal tissue and tumors.

B. Macroscopic NIRAF Screening

• Setup: Mobile Fluorescence Imaging Device.
• Conditions: Tested excitation at 690 nm and 785 nm, detecting emission in the 820–870 nm range.
• Samples: Normal thyroid, parathyroid, pituitary, adrenal glands, and adjacent adipose tissue.

C. In Vivo Intraoperative Study

• Cohort: 27 patients undergoing endoscopic endonasal transsphenoidal surgery for PitNETs.
• Device: Fiber-optic system with a 650 nm laser source (excitation) and an 800 nm long-pass filter (emission). Spectra were acquired using a spectrometer (300–1100 nm range).
• Procedure: A sterile fiber probe was introduced into the surgical corridor under endoscopic guidance. Measurements were taken on:
  1. Normal pituitary tissue.
  2. Tumor tissue.
  3. Other tissues (dura, blood, bone, sinus).
• Validation: Surgeon-reported tissue types were confirmed by postoperative histopathology.
• Statistical Analysis: Receiver Operating Characteristic (ROC) curves with Leave-One-Surgery-Out (LOSO) cross-validation were used to assess discriminative performance.

3. Key Contributions

1. Discovery of Pituitary NIRAF: The study provides the first evidence that the pituitary gland possesses intrinsic Near-Infrared Autofluorescence (NIRAF) sufficient for surgical guidance, similar to the parathyroid gland.
2. Microstructural Identification: Identified secretory granules as the dominant source of long-wavelength NIRAF. Normal pituitary tissue contains a significantly higher density of these granules compared to PitNETs.
3. Label-Free Guidance: Demonstrated that a simple intensity-based measurement (without exogenous dyes) can distinguish normal gland from tumor with near-perfect accuracy.
4. Optimization of Excitation: Validated that red excitation (650 nm) is effective for eliciting pituitary NIRAF, offering a practical alternative to the standard 785 nm excitation used for parathyroid detection.

4. Key Results

Microscopic Findings

• Granule Density: Normal pituitary tissue exhibited a significantly higher area fraction of secretory granules compared to somatotroph and corticotroph PitNETs ($p < 10^{-3}$).
• Spectral Properties: Granules showed a red-shifted emission spectrum (max ~550 nm) with a long-wavelength tail extending to 800 nm. They were excitable by both 405 nm and 640 nm lasers.
• Collagen: Normal tissue also showed higher collagen content than tumors, though granules were the primary driver of the NIR signal.
• Conclusion: The NIRAF contrast is driven by the quantitative abundance of granule-rich microdomains in normal tissue, not a qualitative shift in fluorophore type.

In Vivo Surgical Findings

• Signal Intensity: Normal pituitary tissue consistently showed the highest NIRAF intensity compared to tumors and other tissues (dura, blood, etc.).
• Discriminative Power:
  • Paired Comparison: In every patient where both tissue types were measured, the normal pituitary signal was higher than the tumor signal ($p < 10^{-3}$).
  • ROC Analysis: The method achieved an ROC-AUC of 0.99 (95% CI: 0.98–1.00) for distinguishing pituitary from non-pituitary sites (tumor + other tissues).
  • Separability: There was near-perfect separability between pituitary and non-pituitary measurement sites.
• Signal Dynamics: Continuous probe movement showed step-like transitions in signal intensity, allowing for real-time boundary detection between tissue types.
• Artifacts: Elevated signals in "other tissues" were primarily attributed to the dura overlying the pituitary (proximity effect) or tissue interfaces. Heavy bleeding could suppress the signal.

5. Significance and Clinical Implications

• Enhanced Gland Preservation: This method offers a robust, label-free tool to help surgeons preserve healthy pituitary tissue during tumor resection, potentially reducing postoperative endocrine deficits.
• Workflow Integration: The fiber-optic system is compatible with standard endoscopic endonasal workflows and does not require preoperative injection of contrast agents or changes to surgical timing.
• Mechanistic Insight: By linking the NIRAF signal to secretory granule density, the study provides a biological basis for the technique, suggesting that tumors with lower secretory activity will be easier to distinguish.
• Future Directions: While the current study focuses on distinguishing normal gland from non-gland tissue, future work aims to refine the method to better distinguish between tumor subtypes and adjacent non-gland structures (e.g., cavernous sinus) to aid in complete tumor resection.

Conclusion: The study successfully validates Near-Infrared Autofluorescence (NIRAF) as a highly effective, label-free intraoperative navigation tool for pituitary surgery, capable of distinguishing normal gland tissue from tumors with near-perfect accuracy based on the intrinsic fluorescence of secretory granules.